Abstract

In the first half of the paper, it is argued that while the prospects for a criterion for demarcating scientific theories from pseudoscientific ones are exceedingly dim, it is a mistake to fall back to the position that these differ only with regard to how well they are confirmed. One may admit that different pseudoscientific theories are flawed in different ways yet still insist that their flaws are structural rather than empirical in character. In the second half of the paper, this view is extended to cover the cases of pseudoscientific correlations and therapies, and it is suggested that the pseudosciences are best thought of as radically flawed practices